MIT license ambiguity about copyright?

Question

I asked this question in softwareengineering.stackexchange.com but it was pointed out to me that it is better suited to be posted here. I deleted the original thread in the other subforum.

So the specific question is: Is it allowed to incorporate MIT licensed work in closed-source projects thereby practically bypassing the necessity to deliver with the new work an extra file (or link to a site or something like that) stating a copyright hint of a portion of the code which is licensed under MIT?

Actually the MIT license reads like it is not that ambiguous at all. It states

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

I found these threads on this site, most notably: https://softwareengineering.stackexchange.com/a/264709/262045

even though you legally claimed ownership of the code (as explicitly permitted by the MIT license), you should include a notice to the effect of "portions of this code were originally created by such and such" and state the date of their claim to copyright. But you don't necessarily have to do so since you've re-licensed their project and assumed liability.

but also: https://opensource.stackexchange.com/a/4716/7359 (begs the question is attributing the same as including a copyright notice?)

They read as if I can just just "legally claim ownership" and don't have to state their claim to copyright (doesn't mean I want to strip them of it which I know is not possible by law, just not stating it, the reason being to not give the competition - which is quite strong in my market - a headstart how to just come up with and sell basically the same product. I customized and added to the MIT code heavily, but still the more competitors are out there the higher possibility that at some point someone will be able to reproduce what I made).

On the other hand there are enough sources that state the (for me perceived) different conclusion, like:

https://opensource.stackexchange.com/a/4061/7359

But if you distribute binaries, the license text (typically) gets compiled away, therefore you are no longer fulfilling the single condition in the license. What you need to do is include the license with your binary distributions, maybe as a text file, or embedded in the program's about dialog.

Further complicating the issue, what about if the author of the MIT licensed work answers an email from me saying:

"Yes, the code is free to use (and modify). There is a LICENSE file with MIT license [..describing where to find it..]. A short acknowledge note would be appreciated (but it's not required)."

Again raising the question whether he means attribution or the absolute legal necessity to include the copyright notice.

Most confusing me is the answer by GlenH7 in the above stated thread in https://softwareengineering.stackexchange.com/a/264709/262045 when comparing it to all the evidence to the contrary. He seems credible though and as I cannot comment on his answer not having 50 rep my question is, do I understand him wrong or IS there a indeed quite glaring ambiguity to the MIT license with respect to this matter? And finally, if there is no way to not include the copyright notice under normal circumstances, does the author mailing me these words change that fact in any meaningful and legal way?

Answer 1

even though you legally claimed ownership of the code (as explicitly permitted by the MIT license), you should include a notice to the effect of "portions of this code were originally created by such and such" and state the date of their claim to copyright. But you don't necessarily have to do so since you've re-licensed their project and assumed liability.

The attribution to the original creator is not about liability, it's about recognition. You do necessarily have to retain copyright notices regardless of liability since this is part of the license obligations. If you do not fulfill the license obligations you no longer have a license to distribute the software and could be infringing on someone's copyright.

But if you distribute binaries, the license text (typically) gets compiled away, therefore you are no longer fulfilling the single condition in the license. What you need to do is include the license with your binary distributions, maybe as a text file, or embedded in the program's about dialog.

You also need to include any copyright notices, not just the license text. Regarding where this is best displayed, anything reasonable works. As part of the documentation or in an about box of an application are good options.

"Yes, the code is free to use (and modify). There is a LICENSE file with MIT license [..describing where to find it..]. A short acknowledge note would be appreciated (but it's not required)."

I'm assuming he means that you need to include the LICENSE file (which in turn I assume includes the copyright notices and the MIT license text) but that anything beyond that is not required but nice. While this might suffice, you actually only have to include any copyright statements and the MIT license text. If there is anything beyond that in the LICENSE file, you are free to exclude it. You also do not have to keep this information in a file named LICENSE.

And the original question.

Is it allowed to incorporate MIT licensed work in closed-source projects thereby practically bypassing the necessity to deliver with the new work an extra file (or link to a site or something like that) stating a copyright hint of a portion of the code which is licensed under MIT?

You are allowed to incorporate MIT licensed work in closed-source projects, but you still need to include the copyright notices and the MIT license text in a place where a user of your binary distribution can find it.

Answer 2

As far as I understand (IANAL, IANYL, I don't impersonate one on SE either), the MIT license doesn't allow "... even though you legally claimed ownership of the code (as explicitly permitted by the MIT license),..."

First, the owner of the code is the author(s) (or, if written under contract for somebody else, who hired them). They can transfer their rights to others (in some jurisdictions, the right to be considered the author can not be transferred, however), but just placing the work under the MIT license definitely doesn't allow anybody else to "legally claim ownership". The MIT license allows whoever gets the code a lot of leeway, even use it in closed source software and sell the result for big bucks without giving anything back. Almost the only consideration it asks for is to keep and distribute notices acknowledging the authors/contributors.

Go read the license, it is quite understandable. Check out e.g. the page on the MIT licence at OSI for further details. Copyright law is a complex, convoluted area. And even varies at lot from place to place, today it is under the umbrella of the Berne convention, but even so there are variations.

Answer 3

The MIT license is very permissive, but it does not allow others to "legally claim ownership" in any way, shape, or form. It allows people to use the software, modify it, and even distribute it, but that is not the same as "claiming ownership."

In fact, it would be a violation of the license if someone claimed they created the original code, as the license clearly states that the copyright notice and permission notice must remain intact.

There is no ambiguity about the copyright at all. In fact, the copyright notice is stated in the license.

A person can add their own code, and the code that they create would be copyrighted by them since they created it. But this copyright does not extend to the original code, which is still copyrighted by the original copyright holder.

Where it becomes a mess is when people fail to document which code came from where.

The MIT license is so permissive that leaving the existing copyright and permission notices intact is usually sufficient to comply with the license, but it is better if the source of each code and its version history is documented somehow.